434 PHYSICAL GEOGRAPHY. 



blank strip itself. In fact, this strip, with its end and middle swellings, 

 constitutes a thickened part of the valve, and they have by some writers 

 been called the "median line," and "central" and "terminal nodules." 

 Considerable confusion has arisen in the nomenclature of what might be 

 called the osteology of the diatomaceae. This central thickened band is 

 usually, if not always at some period in the life of the individual, traversed 

 by a canal which runs the whole length of the clear space, but in the 

 thicker ends terminates in enlargements, and is divided into two sections 

 at the centre of the valve, where, likewise, it ends in two round cavities. 

 The end enlargements of this canal have also been called "terminal 

 nodules," and the swellings near the centre have been called "central 

 nodules," as well as the parts just described. Lately it has been proposed 

 to call the tube the "central canal," and by this name we shall designate 

 it in this sketch. At one period in the life of the diatom it would seem 

 that this canal is open outwards down its whole length ; at least, such is 

 the belief of some observers ; but the writer has never been able to satisfy 

 himself that such is the case, for in some of the PinnulaHce he has 

 noticed that the central enlargements of this canal open, by means of 

 trumpet-shaped tubes set at right angles to the course of the canal, into 

 the general cavity of the frustule, and that the terminal expansions, in a 

 like manner, have a communication outwards at the ends of the valve. 

 It is his opinion that this canal has something to do with the motion of 

 the naviculaeform diatoms, which always sail about in a direction parallel 

 to their longest axis. The central canal, when the diatom valve is dead 

 and dry, is filled with air, and then, on account of the effect produced 

 upon the light as it is transmitted through the object to the microscope, 

 appears black, or nearly so, if the objective employed is a good one, and 

 more or less colored when an inferior one is used. The markings found 

 sculpturing the valve of Pinmdaria are different from what we have seen 

 to occur in any of the genera described so far. We find no large hexa- 

 gons nor finer ones here, but, instead, the valve is marked on each side of 

 the median blank space with lines which indicate elevations in the form of 

 bars or corrugations more or less parallel to each other, and set at nearly 

 right angles to the central canal. These bars, or "pinnulae," as they are 

 called, reach from the edge of the valve over the convex margin, and up 

 to the median blank space, where they stop in rounded off extremities. 



